A Phase I trial of 2',3'-dideoxyadenoxine (ddA) was initiated in February of 1988. ddA is a pro-drug of 2'3'-dideoxyinosine (ddI). Whereas ddA is metabolized in the stomach adenine (which can cause renal toxicity). In contrast, ddI is metabolized to hypoxanthine. Thus, ddl appeared to be the preferred form for oral use. With this background, a Phase I trial of ddI was initiated in July of 1988. By July of 1989, it was apparent that: 1) the maximum tolerated dose for long-term therapy was approximately 10 mg/kg/day; 2) doses of 3 to 10 mg/kg were associated with anti HIV activity; 3) dose-limiting toxicities were painful peripheral neuropathy, pancreatitis, and hepatitis; 4) doses of 3 to 10 mg/kg/day were well-tolerated in the majority of patients with AIDS or AIDS-related complex and were associated with long-term clinical and laboratory improvement. Based primarily on the results of this study (with supportive evidence from 2 other Phase I studies), 3 Phase II/III trials of ddI, sponsored by the NIAID and Bristol-Myers Squibb Company were launched in October of 1989 in medical centers around the country. In addition, the FDA enabled patients who could not tolerate AZT or had failed AZT to receive ddI under the mechanisms of a Treatment IND or open label protocol, respectively. At present, more than 10,000 patients have received ddI throughout the United States under these protocols. We are continuing to follow our Phase I patients receiving ddI. We have learned that survival can be excellent with this drug -80% of AIDS patients entered on the study are alive at 20 months. In addition, we have observed that patients with AIDS dementia can have improvement on ddI. Finally, we have observed that patients with extensive prior AZT use have limited CD4 rises on ddI, whereas they do respond with decreases in HIV p24 antigen. We are now exploring the combination of ddI and DHPG (a drug used for retinitis) and ddI used with interferon.